Annual Report
Centre national de la recherche scientifique (CNRS)
Institut de recherche pour le développement (IRD)
27/07/2011 - Article
In the Bani catchment area in Mali, spatial remote sensing has revealed an increase in the vegetation index over the past twenty-five years. This trend, which has been seen in most of the studies done in West Africa, is not accompanied by any significant increases in rainfall. How can we explain it? According to a team from CIRAD, it may be the result of natural vegetation dynamics, rather than of land use changes, as is often suggested.
In the Bani catchment area, which covers 130 000 square kilometres of Mali, a team from CIRAD and its partners monitored vegetation dynamics from 1982 to 2006, to determine what was making the Sahel greener. This was done by looking at a series of NDVI (normalized difference vegetation index) images, which are a good indicator of photosynthetic activity in a given area. The changes in the index were analysed in line with rainfall and land use changes, the main two factors that determine vegetation dynamics in the Sudano-Sahelian zone.
Over the period 1982-2006, there was no change in annual rainfall in the catchment area. However, there was an increase in the areas planted, from 13% to 23% between 1985 and 2000. However, analyses of these results failed to find any clear links between the increased vegetation index in the area and land use changes.
Expanded cropping only explained the trend in northern Bani, in the Sahelian zone, where crops had a higher vegetation index than the natural vegetation. The increased index in the catchment area as a whole was thus primarily due to natural vegetation dynamics.
This dynamic in fact depended on rainfall distribution over the past twenty-five years, and not on rainfall trends. In effect, while there was a rainfall deficit from 2000 to 2006, rainfall increased from 1982 to 1999. Perennial plants were able to make use of that increase and survive during the subsequent dry period.
The study showed the limitations of trend analyses in remote sensing and climatology, based on linear relations, which are too simplistic to fully reproduce the ecological and geographical phenomena at play.
Bégué A. et al., 2011. Can a 25-year trend in vegetation dynamics (NOAA-AVHRR NDVI) be interpreted in terms of land use change? A case study of the Bani catchment in Mali. Global Environmental Change, 21: 413-420. Doi : 10.1016/j. gloenvcha.2011. 02.002.